Message-ID:  <E0wIAwh-0004xk-00@heaton.cl.cam.ac.uk>
Reply-To: Mailing list for the LaTeX3 project
              <LATEX-L@RELAY.URZ.UNI-HEIDELBERG.DE>
In-Reply-To:  Your message of "Fri, 18 Apr 1997 12:08:27 +0200." 
              <v02130504af7cf2e152c9@[130.237.37.66]>
Date:         Fri, 18 Apr 1997 11:31:10 +0100
From: Robin Fairbairns <Robin.Fairbairns@CL.CAM.AC.UK>
Sender: Mailing list for the LaTeX3 project
              <LATEX-L@RELAY.URZ.UNI-HEIDELBERG.DE>
To: Multiple recipients of list LATEX-L
              <LATEX-L@RELAY.URZ.UNI-HEIDELBERG.DE>
Subject:      Re: Alternatives to LaTeX
Status: R

Hans Aberg writes:

> [Much of this has already been answered by Robin Fairbairns]

I think not.

> > > One can convert PS font metric to TeX font metric, but TeX
> > > can have more glyphs in a font, which allows for more kerning information,
> > > for example.
> >this is, excuse me, a meaningless sentence. TeX can see no more glyphs
> >in a font than it has...
>
>   This is really for the experts to reply, but most implementations of PS
> fonts do not use several of the 256 possible slots, but this would not be
> efficient for TeX fonts, which would can use all those slots. In addition,
> you can only do kerning for glyph pairs in the same font of 256 glyphs. So
> for building a good TeX font, you would have to build a virtual font based
> on several PS fonts, adding new kerning information.

Suppose you have a type1 font that offers all the glyphs you need.
What is the (efficiency) advantage to making a composite virtual font,
rather than simply using what you've got?  In my experience (almost
exclusively typesetting Latin-script text) the automatically-generated
8r encoding of type1 gives me all I need.

In such a case, it's a ludicrous *waste* of effort to create a new
virtual font: unless you're wandering into typesetting of languages
that aren't covered by 8r, to do such a thing would be ridiculously
*in*efficient.

> > >   So the problem is that it is a lot of work setting it up.
> >
> >only once
>
>   As Robin Fairbairns pointed out, there is a lot of work involved in
> adding full METAFONTness, but you can of course also ignore adding that
> missing information.

Sebastian was answering a different question.  You were talking (in a
separate branch of the thread) about the awfulness of typesetting
using linearly-scaled type1 fonts, and Sebastian retorted that if it's
what the designer specifies, it's what he'll do.

Michael Downes remarked that linear scaling is bad when the sizes get
really small, which only really happens in maths; you then responded
that we need a "bit of work" to add meta-ness to the existing fonts,
and I responded to suggest that the "bit" was more likely to involve
armies of designers and metafont hackers.

>   In fact there seems to be tw owholly different discussion topics going on
> here:

I.e., you actually _knew_ all the above...

Robin (loudmouth) Fairbairns